Abstract: Government information resources retrieval is important function in directory service system.Based on the XML metadata standard stipulated in national standards of Government information resource directory system，a keywords search algorithm is proposed，which uses the XML TF*IDF ranking strategy of government information resource metadata and the keywords dependence to rank the individual matches by semantic relevance.An improved keywords inverted index is proposed to improve the query efficiency.The experimental results show that this algorithm can greatly improve the rank accuracy of search results as well as the time efficiency，which can effectively improve the data-sharing ability of government information resource.

Key words: government information resource, metadata, keyword search, semantic relevance, Extensible Markup Language（XML）

摘要： 政务信息资源检索是政务信息资源共享系统的重要功能。以《政务信息资源目录体系》国家标准中的XML元数据规范为依据，提出了一种支持关键词搜索的政务信息资源检索算法。该算法使用政务信息资源XML元数据的TF*IDF和关键词依赖度对检索结果集进行语义相关度排序，通过改进关键词倒排索引来提高检索效率。实验表明该算法在检索结果排序精确度和时间效率上均有较大的改善，可有效提高政务信息资源利用的数据共享服务能力。

关键词: 政务信息资源, 元数据, 关键词检索, 语义相关度, 可扩展标记语言（XML）